Applied ML Engineer: Research to Production (Italy)

Lever, Inc.; Italia; In loco; EUR 34.000 - 45.000; Tempo pieno

Descrizione del lavoro

Lever, Inc. is seeking an Applied ML Engineer based in Italy to bridge ML research and production systems. You will turn research ideas into repeatable experiments, build evaluation tooling, and contribute across research, experimentation, engineering, and product.

The role emphasizes ownership, rigor, and cross-stack collaboration, with a focus on ML provenance and verification. The position involves designing experiments, handling model internals, and delivering product-ready interfaces and

Competenze

• Strong Python engineering skills with hands-on PyTorch experience.
• Proficient in ML evaluation, dataset design, baselines and metrics.
• Able to read ML research papers and implement methods from first principles.
• Professional software engineering experience beyond notebooks (APIs, async jobs, databases, logging, tests).
• Familiarity with open-weight models and LLM inference systems.

Mansioni

• Reproduce and evaluate ML research methods using open-weight models.
• Design evaluation datasets, probes, baselines, and experiment harnesses.
• Work with model weights, logits, hidden states, APIs, and inference infra.
• Build evaluation infrastructure for runners, judges, persistence, orchestration, and reporting.
• Turn research workflows into intuitive product experiences and dashboards.

Conoscenze

Python

PyTorch

Hugging Face Transformers

React/TypeScript

Experimentation

Strumenti

DSPy

LiteLLM

Temporal

Ray

PostgreSQL/pgvector

vLLM

Descrizione del lavoro

Lever, Inc. is seeking an Applied ML Engineer based in Italy to bridge ML research and production systems. You will turn research ideas into repeatable experiments, build evaluation tooling, and contribute across research, experimentation, engineering, and product.

The role emphasizes ownership, rigor, and cross-stack collaboration, with a focus on ML provenance and verification. The position involves designing experiments, handling model internals, and delivering product-ready interfaces and